Cancer survival analyses based on
cancer registry data do not provide direct information on the main aim of
cancer treatment, the cure of the patient. In fact, classic survival indicators do not distinguish between patients who are cured, and patients who will die of their disease and in whom prolongation of survival is the main objective of treatment. In this study, we applied parametric cure models to the
cancer incidence and follow-up data provided by 49 EUROCARE-4 (European
Cancer Registry-based study, fourth edition)
cancer registries, with the aims of providing additional insights into the survival of European
cancer patients diagnosed from 1988 to 1999, and of investigating between-population differences. Between-country estimates the proportion of cured patients varied from about 4-13% for
lung cancer, from 9% to 30% for
stomach cancer, from 25% to 49% for colon and
rectum cancer, and from 55% to 73% for
breast cancer. For all
cancers combined, estimates varied between 21% and 47% in men, and 38% and 59% in women and were influenced by the distribution of cases by
cancer site. Countries with high proportions of cured and long fatal case survival times for all
cancers combined were characterised by generally favourable case mix. For the European pool of cases both the proportion of cured and the survival time of fatal cases were associated with age, and increased from the early to the latest diagnosis period. The increases over time in the proportions of Europeans estimated cured of lung, stomach and colon and
rectum cancers are noteworthy and suggest genuine progress in
cancer control. The proportion of cured of all
cancers combined is a useful general
indicator of
cancer control as it reflects progress in diagnosis and treatment, as well as success in the prevention of rapidly fatal
cancers.